Combining Exact and Metaheuristic Techniques for Learning Extended Finite-State Machines from Test Scenarios and Temporal Properties

Chivilikhin, Daniil; Ulyantsev, Vladimir; Shalyto, Anatoly

doi:10.1109/icmla.2014.62

Cited by 5 publications

(21 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This subsection reports on the comparison of the proposed FSM identification techniques with three inexact methods: the CSP+MuACOsm algorithm [9], the state merging approach [38], and the Unbeast tool [16] based on bounded LTL synthesis. Since the performance of the QSAT-based method has been shown to be inappropriate, this method was excluded from experiments in this subsection.…”

Section: Comparison With Inexact Methodsmentioning

confidence: 99%

“…This time, since CSP+MuACOsm is not an exact method, we did not try to minimize |S| and executed the methods for the maximum possible numbers of states. Note that [9] used the AMD Phenom II x 4 955 3.2 GHz CPU for experiments. As for the proposed methods, which were still executed on an Intel Core i 7-4510U 2.0 GHz CPU, they were now given 15 minutes for l " 50|S|, 30 minutes for l " 100|S| and 60 minutes for l " 200|S|, since solver execution time was affected by the length of scenarios Table 6 Median execution times (in seconds) of the proposed methods and CSP+MuACOsm (designated as CMA) on the instance sets from [9].…”

Section: Comparison With the Metaheuristic Fsm Identification Methodsmentioning

confidence: 99%

“…They are further compared with existing inexact approaches. First, two of them are shown to outperform the approach from [9]. Compared to state merging [38], the proposed approaches need more time, but are applicable under fewer premises (such as the absence of actions on transitions).…”

Section: Introductionmentioning

confidence: 99%

“…Recently, there has been growing interest in automated FSM construction based on given specifications, which are often represented as execution traces and logs [23,38,39,31]. Other types of data employed in model construction are temporal properties [38,9,31] and invariants [3]. This research direction is appealing, since inferred finite-state models can help comprehend software, reveal faults in it, facilitate model-driven development, or even serve as software.…”

Section: Introductionmentioning

confidence: 99%

“…Existing techniques, such as state merging [27,38] and metaheuristic approaches [34,9], demonstrate acceptable performance. However, they are almost not concerned about the size of generated models: it is not always possible to obtain the FSM with the minimum number of states, and even when it is, existing methods do not provide the proof that the found automaton is indeed the smallest possible one.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Exact finite-state machine identification from scenarios and temporal properties

Ulyantsev

Buzhinsky

Shalyto

2016

Int J Softw Tools Technol Transfer

Self Cite

View full text Add to dashboard Cite

Finite-state models, such as finite-state machines (FSMs), aid software engineering in many ways. They are often used in formal verification and also can serve as visual software models. The latter application is associated with the problems of software synthesis and automatic derivation of software models from specification. Smaller synthesized models are more general and are easier to comprehend, yet the problem of minimum FSM identification has received little attention in previous research. This paper presents four exact methods to tackle the problem of minimum FSM identification from a set of test scenarios and a temporal specification represented in linear temporal logic. The methods are implemented as an open-source tool. Three of them are based on translations of the FSM identification prob- lem to SAT or QSAT problem instances. Accounting for temporal properties is done via counterexample prohibition. Counterexamples are either obtained from previously identified FSMs, or based on bounded model checking. The fourth method uses backtracking. The proposed methods are evaluated on several case studies and on a larger number of randomly generated instances of increasing complexity. The results show that the Iterative SAT-based method is the leader among the proposed methods. The methods are also compared with existing inexact approaches, i.e. the ones which do not necessarily identify the minimum FSM, and these comparisons show encouraging results.

show abstract

Section: Comparison With Inexact Methodsmentioning

confidence: 99%

Section: Comparison With the Metaheuristic Fsm Identification Methodsmentioning

confidence: 99%